Lead Data Scientist Offline

Primary Responsibilities

• Take full ownership in developing effective ways to do various analyses like customer segmentation, user modeling, churn analysis and prediction, LTV prediction, and otherwise process large volumes of customer and device data with the intention of gaining actionable insights and making data-driven decisions.

• Collaborate closely with other team members including data scientists, engineers, and quality analysts to make sure the team is progressing as a cohesive unit and producing results. Provide technical guidance to other team members when required.

• Help multiple audiences - engineers, product managers, executives - dig into and understand the output of the team’s analyses and models using notebooks (e.g. Rmd, Jupyter), visualization, presentations etc.

• Constantly look for and adopt new techniques and tools to ensure the team stays at the forefront of modern large-scale data processing and analysis techniques

Skills - Experience and Requirements

If you meet most of the following requirements, you are likely to be a great fit for the position:

• You have an academic background in applying statistics and machine learning. The typical candidate has a Bachelor’s or Master’s degree in Math, Statistics, Computer Science, or Physics or such quantitative fields or has done a program from a business school in marketing, analytics etc. with a focus on quantitative approaches.

• You have at least 5 years of experience working with data and data analysis in some form. You have built predictive models and have done analyses to explain and understand the underlying process, variable relationships, causality etc.

• You have a wide range of statistical and machine learning tools under your belt, and deep practical insight to choose the best tools for a given problem. These include linear models for regression and classification, multi-level models, factor analysis & PCA, discriminant analysis, support vector machine, decision tree ensembles & bootstrap, neural networks, mixture models & clustering algorithms, and so on.

• You know how to incorporate prior domain knowledge in your models in a principled fashion. You are familiar with Bayesian modeling and inference; you also know when and when not to use them based on practical considerations.

• You are proficient in at least one programming language commonly used for data analysis (like R/Python), and you are comfortable with SQL

Additional Preferred qualifications

• Experienced in designing and analyzing controlled experiments. This will be a strong plus.

• Possess strong data visualization skills using programmatic tools (e.g. ggplot2, shiny, d3.js) and other visualization frameworks like victory, highcharts etc. This will be a strong plus.

• Have worked on problems involving survival analysis or time series modeling

• Have worked with large data sets, with big data processing tools like MapReduce, Spark, Hive, etc. Have data engineering skills to do basic preprocessing, cleaning and transformations.

• Knowledgeable on different database and data warehousing systems like MySQL, Amazon Redshift, BigQuery, Teradata